Essbase unix file system best practice

Is there such thing in essbase as storing files in different file system to avoid i/o contention? Like for example in Oracle, it is best practice to store index files and data files indifferent location to avoid i/o contention. If everything in essbase server is stored under one file directory structure as it is now, then the unix team is afraid that there may run into performance issue. Can you please share your thought?
Thanks

In an environment with many users (200+) or those with planning apps where users can run large long-running rules I would recommend you separate the application on separate volume groups if possible, each volume group having multiple spindles available.
The alternative to planning for load up front would be to analyze the load during peak times -- although I've had mixed results in getting the server/disk SME's to assist in these kind of efforts.
Some more advanced things to worry about is on journaling filesystems where they share a common cache for all disks within a VG.
Regards,
-John

Similar Messages

Uploaded Files stored in Oracle 10G database or in Unix File system

Hey All,
I am trying to understand best practices on storing uploaded files. Should you store within the database itself (this is the current method we are using by leveraging BLOB storage) or use a BFILE locator to use the files system storage (we have our DB's on UNIX) . . .or is there another method I should be entertaining? I have read arguments on both sides of this question. I wanted to see what answers forum readers could provide!! I understand there are quite a few factors but the situation I am in is as follows:
1) Storing text and pdf documents.
2) File sizes range from a few Kb to up to 15MB in size
3) uploaded files can be deleted and updated / replaced quite frequently
Right now we have an Oracle stored procedure that is uploading the files binary data into a BLOB column on our table. We have no real "performance" problems with this method but are entertaining the idea of using the UNIX file system for storage instead of the database.
Thanks for the insight!!
Anthony Roeder

Anthony,
First word you must learn here in this forum is RESPECT.
If you require any further explanation, just say so.
BLOB compared with BFILE
Security:
BFILEs are inherently insecure, as insecure as your operating system (OS).
Features:
BFILEs are not writable from typical database APIs whereas BLOBs are.
One of the most important features is that BLOBs can participate in transactions and are recoverable. Not so for BFILEs.
Performance:
Roughly the same.
Upping the size of your buffer cache can make a BIG improvement in BLOB performance.
BLOBs can be configured to exist in Oracle's cache which should make repeated/multiple reads faster.
Piece wise/non-sequential access of a BLOB is known to be faster than a that of a BFILE.
Manageability:
Only the BFILE locator is stored in an Oracle BACKUP. One needs to do a separate backup to save the OS file that the BFILE locator points to. The BLOB data is backed up along with the rest of the database data.
Storage:
The amount of table space required to store file data in a BLOB will be larger than that of the file itself due to LOB index which is the reason for better BLOB performance for piece wise random access of the BLOB value.

Download PDF spool to unix file system

Does anyone know how to programatically download a PDF document spool to the unix file system?
I am trying to find a method to send PDF documents that have gone to spool to a unix file system. If anyone had any ideas on how to do this, please let me know.
Thanks.

Hi,
For this define a logical file while using transaction FILE.
In your code you will get the complete path of the file while using FM FILE_GET_NAME.
Then download your PDF while using instruction
OPEN DATASET...
TRANSFERT ...
Hope this help you .
Best regards

External Tables to Unix File System 10G R2

Can anyone help with setting up an external table that reads a flat file from a Unix File system.
I have sampled a file ok and created an external table and deployed it to the database ok but it can find the link through to the unix file system to read the file.
I created the location as an FTP type and have referenced the path to the relevant directory /oracle02/app/OWB_files when creating the location. I have placed the relevant named file in the directory but when i try to look at the table in TOAD i get the following errors
ORA-29913 error in executing ODCIEXTTABLEOPEN callout
ORA-29400 data cartridge error
KUP-04040 files 121123_PENS.txt in MLCC_FILES_LOCATION_0 not found
Does anyone have a step by step guide for creating these and am i doing some thing wrong. Is choosing the FTP type in the location correct and is the path specified correctly. I can see much information on thisin the manual!
Your assistance would be appreciated

HI,
You make sure that, the path should be shared one. We can do this using samba server.
Regards,
Gowtham Sen.

IPod formatted to UNIX File System (UFS)

Hello all, I had an old 15G iPod (the generation with the 4 button above the wheel) and I decided to reformat it and use it as a firewire drive, it's not used for music anymore. Anyway, I was about half way through the format and I noticed I had accidently sellected UNIX File System (UFS) instead of Mac OS Extended. Well I let lt finish and thought I'd just reformat it again to Mac OS Extended, only now the iPods not even being recognized by OSX? I looked in Disk Utility and nothing there either. iPod restore does nothing either, just says connect an iPod. There's an Apple on the screen and that's it? What gives? Will I be able to get this thing running again?

Thanks but no thanks! I got it going using info at this link:
http://ipodlinux.org/Installation
Got it in Disk Mode!!!

Open API App (Windows based) fails to open FMB on Unix file system

My Open API, Windows based app, can successfully open and 'get' properties of FMBs stored in the Windows file system. However, it fails to load the FMB when the FMB resides on a networked Unix server. The same FMBs on Unix can be opened by the Windows based FormBuilder (over the network). I can copy the FMB down to Windows and without re-compiling the FMB, my Open API app can 'load' the FMB and 'get' all the properties. What suggestions can you give for debugging / resolving this? I need to be able to 'Load' the FMBs (through the Open API), that reside in the Unix file system, from Windows.
JJ

Generally, this is why we will tell you that accessing net shares is not supported and in places where it might even be supported, we would still suggest that it is not recommended. Accessing via net shares (especially through Windows) is often problematic. There are various performance and connectivity issues that, unfortunately fool you into believing that the product you are using is flawed when the problem is really a connection issue with the share.
In your case, because you are not exactly using an Oracle product (initially), Oracle can't offer much anyway, but I would recommend against using shares whenever possible. If you need to access a file, copy it locally first, perform whatever task on it, then return the updated file to its origin. This method protects you from things like net failure and instability as well as the performance issues associated with accessing files remotely.

UNIX File System?

I was able in the past to format a disk with UNIX File System and read and write to it using Disk Utility in OS X Tiger. Now, my new Leopard machine does not show that option. What happened to that? How do I get it back?
Thanks,
John

I'm pretty sure UFS support is totally gonzo in leopard. you certainly [can not install Leopard|http://docs.info.apple.com/article.html?artnum=306516] on UFS formatted drives.

Java script in HTMLDB to check if file exists in Unix file system

How do I use javascript to check if file is exists in Unix file system. I would like to dispaly the columns only if file is exists.

Hello,
This is one of those features that the manuals do not cover.
How to use and build AJAX features could be a whole book all by itself, and it's not really HTML DB specific feature even though we have built some hooks in application and javascript to make it easier.
Take a look at this thread
Netflix: Nice UI ideas
and I've built some examples here
http://htmldb.oracle.com/pls/otn/f?p=11933:11
Or just search the forums for AJAX or XMLHTTP
Carl

Unix file system & Oracle datafiles--urgent plz

How i can chech my oracle db files on which unix file system? In HP/UX exvirnment??

select * from dba_data_files
AUTOEXTENSIBLE column gives you whether autoextend is on or not.
Join with dba_free_space to get free space for each file.
You can check the following link
http://www.oracle.com/technology/oramag/code/tips2003/083103.html

File import best practice

I need some outside input on a process. We get a file from a bank and I have to take it and move it along to where it needs to go. Pretty straight forward.
The complexity is the import process from the bank. It's a demand pull process where an exe needs to be written that pulls the file from the bank and drops it into a folder. My issue is they want me to kick the exe off from inside SSIS and then use a file
watcher to import the file into a database once the download is complete. My opinion is the the SSIS package that imports the file and the exe that gets the file from the bank should be totally divorced from each other.
Does anybody have an opinion on the best practice of how this should be done?

Here it is:http://social.msdn.microsoft.com/Forums/sqlserver/en-US/bd08236e-0714-4b8f-995f-f614cda89834/automatic-project-execution?forum=sqlintegrationservices
Arthur My Blog

Application preferences system - best practice

Hi,
I'm migrating a forms application from 6i to 11g
The application source forms are the same for all the possible deployments , which are set in different databases (let's say one database for each client).
But as you may expect, each client wants a customized app, so we have to define some kind of preferences, and have them into account in forms (and db packages).
The problem:
The application, as it was designed, has this customizing system spread over different solutions:
A database table with one row for each custom parameter.
A database package constants.
Forms global variables defined at the main menu.
Even, instead of defininig a good set of properties, I'm finding a lot of code with "if the client is one of this then ... else ..." sentences. Sometimes implemented with instr and global variables defining groups of clients ... bufff....
The question:
I'd like to take advance of the migration process to fix this a little bit. Can you give advice on a best practice for this?

Thanks. I was hoping there would be something better than both approaches (package constants or parameter table) bundled within the database.
Of course the
if customer name = 'COMPANY_A' then use this logic ...
gives the creeps to anyone.
Instead of this everything should be
if logic_to_do_this = 'A' then ...
There are two minor problems with the table approach:
Single row and one column for each parameter? (thus you can control the parameter datatype, but need to ddl everytime you add a parameter.
or a parameter table with parameter name/value pairs? (here you'd have to go with varchar for everything, but you could have a set of pre-established conversion masks)
I prefer the second (you only need to establish masks for number and date parameters), but I'm a bit late, as the app has been working with a single row parameter table from the beginning, and they even didn't wrap it with a getter function.
In fact, in an old forms application I developed where customization was paramount, I remember I used two tables master / detail: the master table for the parameter "context" definition and the detail table for the values and dates of entry in force (the app worked 24x7 and users even needed to program changes in preferences in advance). A getter function gets you the value currently in force.

Flat File load best practice

Hi,
I'm looking for a Flat File best practice for data loading.
The need is to load a flat fle data into BI 7. The flat file structure has been standardized, but contains 4 slightly different flavors of data. Thus, some fields may be empty while others are mandatory. The idea is to have separate cubes at the end of the data flow.
Onto the loading of said file:
Is it best to load all data flavors into 1 PSA and then separate into 4 specific DSOs based on data type?
Or should data be separated into separate file loads as early as PSA? So, have 4 DSources/PSAs and have separate flows from there-on up to cube?
I guess pros/cons may come down to where the maintenance falls: separate files vs separate PSA/DSOs...??
Appreciate any suggestions/advice.
Thanks,
Gregg

I'm not sure if there is any best practise for this scenario (Or may be there is one). As this is more data related to a specific customer needs. But if I were you, I would handle one file into PSA and source the data according to its respective ODS. As that would give me more flexibility within BI to manipulate the data as needed without having to involve business for 4 different files (chances are that they will get them wrong - splitting the files). So in case of any issue, your trouble shooting would start from PSA rather than going thru the file (very painful and frustating) to see which records in the file screwed up the report. I'm more comfortable handling BI objects rather than data files - coz you know where exactly you have look.

FDM file format best practice

All, We are beginning to implement an Oracle GL and I have been asked to provide input as to the file format provided from the ledger to process through FDM (I know, processing directly into HFM is out..at least for now..).
Is there a "Best Practice" for file formats to load through FDM into HFM. I'm really looking for efficiency (fastest to load, easiest to maintain, etc..)
Yes, we will have to use maps in FDM, so that is part of the consideration.
Questions: Fix or delimited, concatenate fields or not, security, minimize the use of scripts, Is it better to have the GL consolidate etc...?
Thoughts appreciated
Edited by: Wtrdev on Mar 14, 2013 10:02 AM

If possible a Comma or Semi-Colon Delimited File would be easy to maintain and easy to load.
The less use of scripting on the file, the better import performance.

DW:101 Question - Site folder and file naming - best practices

OK - My 1st post! I’m new to DW and fairly new to developing websites (have done a couple in FrontPage and a couple in SiteGrinder), Although not new at all to technical concepts building PCs, figuring out etc.
For websites, I know I have a lot to learn and I'll do my best to look for answers, RTFM and all that before I post. I even purchased a few months of access to lynda.com for technical reference.
So no more introduction. I did some research (and I kind of already knew) that for file names and folder names: no spaces, just dashes or underscores, don’t start with a number, keep the names short, so special characters.
I’ve noticed in some of the example sites in the training I’m looking at that some folders start with an underscore and some don’t. And some start with a capital letter and some don’t.
So the question is - what is the best practice for naming files – and especially folders. And that’s the best way to organize the files in the folders? For example, all the .css files in a folder called ‘css’ or ‘_css’.
While I’m asking, are there any other things along the lines of just starting out I should be looking at? (If this is way to general a question, I understand).
Thanks…
\Dave
www.beacondigitalvideo.com
By the way I built this site from a template – (modified quite a bit) in 2004 with FrontPage. I know it needs a re-design but I have to say, we get about 80% of our video conversion business from this site.

So the question is - what is the best practice for naming files – and especially folders. And that’s the best way to organize the files in the folders? For example, all the .css files in a folder called ‘css’ or ‘_css’.
For me, best practice is always the nomenclature and structure that makes most sense to you, your way of thinking and your workflow.
Logical and hierarchical always helps me.
Beyond that:
Some seem to use _css rather than css because (I guess) those file/folder names rise to the top in an alphabetical sort. Or perhaos they're used to that from a programming environment.
Some use CamelCase, some use all lowercase or special_characters to separate words.
Some work with CMSes or in team environments which have agreed schemes.

Integrating Multiple systems - best practice

Hi,
I need to integrate the following scenario let me know the best practice with steps.
Scenario is end to end syncrhonous
A system(supports open tech i.e webservice client) => SAP PI<=>Oracle(call stored procedure) => Update in ECC=> Reponse back to A system (source)
Thanks
My3

Hi Mythree,
First get the request from the web service to pi and then map to stored procedure using synchronous send step (Sync Send1) in bpm and get the response back. Once when you get the response back from oracle then use another synchronus send step. Here take the response from the database and map to ecc (use either a proxy or rfc) which is Sync Send2 in the bpm and get the udpdates back. Once when you get the response back send back the response to source system. The steps in BPM would be like this:
Start --> Receive --> Sync Send1 --> Sync Send2 --> Send --> Stop
These blogs might be useful for this integration:
/people/siva.maranani/blog/2005/05/21/jdbc-stored-procedures
/people/luis.melgar/blog/2008/05/13/synchronous-soap-to-jdbc--end-to-end-walkthrough
Regards,
---Satish

Essbase unix file system best practice

Similar Messages

Maybe you are looking for